For recent years, our group has specialized in finding correlations between the photovoltaic parameters of perovskite solar cells (PSCs) or dye-sensitized solar cells (DSSCs) and the results of time resolved spectroscopic experiments (especially on the ultrafast and fast time scales). In particular, we have monitored the decay of the excited population in active material (perovskite or dye) by means of emission and transient absorption on the time scales from femtoseconds to nanoseconds. An interesting question is whether the faster observed decay is accompanied with more efficient charge separation (and thus higher photocurrent) or not.

On the one hand, faster decay of the transient signals can be due to the more rapid charge injection from perovskite or dye to the contact charge transporting materials. In this case the faster process often means higher photocurrent of the solar cell. We have observed such correlation for PSCs prepared in our lab in Poznań: in recent studies of triple cation perovskite on titania layers of different quality, and in our previous studies on different synthesis conditions for double cation perovskite [1]. For DSSC, we have observed that slower electron injection from carbazole dyes to TiO₂ results in smaller relative photocurrent of the cells e.g. in water-based electrolyte [2] or after applying molecular or atomic passivation layers on dye-titania interface [3]. However, too fast charge injection can be accompanied with too fast back electron transfer which lowers the photocurrent. It was e.g. the case of carbazole dyes without co-adsorbent [3], or without tert-butylpyridine in electrolyte [4]. The optimum balance between forward and back electron transfer was observed for injection time constant 0.5-1 ps [3]. From our earlier PSC studies, the hole injection was faster from standard MAPbI₃ to CuSCN than to spiro-OMeTAD, but the photocurrent was higher with the latter material due to slower charge recombination at the interface [5].

On the other hand, faster decay of the transient signals can be caused by the unwanted deactivation processes, competing with charge injection. For DSSC, we have observed the role of self-quenching in indoline dyes: slower decay of the excited state upon adding co-adsorbent [6] or molecular capping [7] resulted in higher photocurrents. For inverted PSC, the first order decay rate constant was higher for the samples not covered by any electron transporting material than that in full cells, probably due to increased trap-assisted recombination at the defects on perovskite outer surface [8].